THE JOURNAL OF BIOLOGICAL CHEMISTRY

Printedin U.S.A. Vol. 257, No. 1. Issue of January 10, pp, 198-207, 1982

Changes in the Distribution and Composition of Plasma High Density Lipoproteins after Ingestion of Fat*

(Received for publication, June 15, 1981)

Alan R. Tall$, Conrad B. Blum, Gary P. Forester, and Charles A. Nelson From the Department of Medicine, Columbia University, College of Physicians and Surgeons, New York, New York 10032 and the Department of Biochemistry, University of Arkansas, School of Medical Sciences, Little Rock, Arkansas 77205

Following ingestion of a fatty meal there is an in- crease in concentration of phospholipids and proteins in the plasma high density lipoproteins (HDL). To eval- uate the resulting changes in HDL subclasses, the plasma HDL of six subjects were analyzed 4 to 8 h after ingestion of 100 ml of corn oil or 80 ml of corn oil with four eggs. Isopycnic density gradient ultracentrifuga- tion of fasting plasma showed two broad components of HDL: a major peak of density (d) 1.11 to 1.17 g/ml (HDL3) and a smaller peak of d 1.07 to 1.11 g/ml (HDL2). Following ingestion of either type of fatty meal, there was an increase in lipoprotein mass in both peaks of HDL and their centers of mass were shifted to lower density (1.140 + 1.120 to 1.130 g/ml; 1.095 + 1.090 g/ ml). Calculation of changes in HDL concentration (li- pemic minus fasting) showed that the alterations in density gradient profile were due to a major increase in lipoproteins of d 1.102 to 1.137 g/ml, a smaller in- crease in a separate lipoprotein peak of 1.080 to 1.102 g/ml, and a small decrease in lipoproteins of d 1.137 to 1.165 g/ml. Redistribution of HDL mass into larger, less dense lipoproteins was also demonstrated by agarose gel chromatography or by minimal spin density gra- dient ultracentrifugation in a vertical rotor.

The increase in mass of 1.080 to 1.102 lipoproteins was largely due to increased concentrations of phos- pholipid, cholesterol ester, and apoA-I, while the in- crease in 1.102 to 1.137 lipoproteins was due to in- creased concentrations of apoA-I, apoA-11, phospholip- ids, cholesterol, and cholesterol esters. Analytical ul- tracentrifugation of representative samples within these density intervals showed lipoprotein species with molecular weights and sedimentation coefficients, re- spectively, of 378,000, 5.8 (d 1.080 to 1.095); 248,000, 3.5 (d 1.110 to 1.120); and 173,000, 1.6 (d 1.135 to 1.150). Polyacrylamide gradient gel electrophoresis showed that the 1.080 to 1.102 lipoproteins contained a single lipoprotein band of diameter -10.7 nm; the 1.102 to 1.137 lipoproteins contained a single band which varied in size from 10.0 to 9.2 nm: and the 1.137 to 1.165 lipoproteins contained three species of diameters -9.2, 8.8, and 8.2 nm. Within density intervals, the molecular weights, sedimentation coefficients, and diameters of the different lipoproteins were similar in fasting and lipemic plasma. Calculation of average molecular com- positions shows that the major incremental HDL of d -

* This research was supported by grants HL 22682, HL 21006, and T 32-AM 07330 from the National Institutes of Health and by a grant- in-aid from the American Heart Association (316-3070-2286, New York affiliate). The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore be hereby marked “advertisement” in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.

f An Established Investigator of the American Heart Association.

1.12 g/ml could be derived by addition of lipids to the largest species of fasting HDL3. Within density inter- vals, the particle contents of apoA-I and apoA-11 were unchanged during lipemia, suggesting that apoprotein transfer causes interconversion of existing HDL species or formation of new particles with the same content of apoA-I and apoA-II as existing species.

Several population studies have shown a strong inverse correlation between plasma HDL’ levels and incidence of atherosclerotic cardiovascular disease (1, 2). Although there is some evidence that HDL may remove cholesterol from the tissues (3), the reason for their association with protection against atherosclerosis remains uncertain. The epidemiologi- cal observations have stimulated investigations of HDL struc- ture and metabolism. HDL or its components are derived from a variety of secretory and lipolytic sources. HDL appears to be secreted as a nascent discoidal particle by the rat liver (4) and small intestine (5). Such discoidal particles may be converted into spherical HDL as a result of cholesterol ester formation by plasma lecithin: cholesterol acyltransferase (6). HDL components are also derived by transfer from chylomi- crons or very low density lipoprotein during lipolysis (7-13). Thus, a portion of very low density lipoprotein phospholipids, cholesterol, and apoprotein C are transferred into HDL during catabolism in the perfused rat heart (12). Also, intravenous injection of intestinal lymph chylomicrons into rats results in transfer of phospholipids (7, 13) and apoA-I (7) into HDL. There is an increase in HDL phospholipids and protein during alimentary lipemia in man (14, 15), suggesting that transfer of chylomicron surface constituents into HDL may also occur in humans.

The plasma HDL are heterogeneous (16-19). Based on the schlieren profile of HDL in the analytical ultracentrifuge, HDL were originally classified into two subclasses, HDL, and HDL3, which can be isolated at densities 1.063 to 1.125 and 1.125 to 1.21 g/ml in the preparative ultracentrifuge. More recently, isopycnic density gradient ultracentrifugation (18, 19) and polyacrylamide gradient gel electrophoresis (18, 20, 21) have suggested further subclassification of HDL into three major subclasses: HDL, (1.125 to 1.210 g/ml), HDL2, (1.100 to 1.125 g/ml), and HDL*I, (1.063 to 1.100 g/ml). Further subclasses within HDL3 have been described (16,20,21). Also metabolically important subfractions have been isolated from HDL by affinity chromatography (17, 22).

A major unanswered question is: how do the different secretory or lipolytic sources of HDL influence the composi- tion and structure of subclasses of plasma HDL? As an ap-

’ The abbreviations used are: HDL, high density lipoprotein; apoA- I and apoA-11, apolipoproteins A-I and A-11.

198

by guest on January 3, 2020http://w

ww

.jbc.org/D

ownloaded from

HDL Subclasses after Fat Intake 199

TABLE I Plasma lipids and apoproteins after fat ingestion

Mean f S.E. changes (milligrams per dl plasma) in plasma triglyceride (TG), cholesterol (CHOL), apolipoprotein A-I, and apolipoprotein A-I1 at different times after fat intake.

Time after corn oil Time after corn oil with eggs ~~ ~

Oh 4 h 5 h 6 h 8 h Oh 4 h 5 h 6 h 8 h

na 6 4 5 6 4 6 4 5

CHOL 1 2 9 k 3 1 3 3 f 3 1 3 1 f 3 1 3 1 f 5 1 3 7 f 4 1 2 7 f 3 1 2 3 k 5 1 2 7 f 8 1 3 1 f 2 131-t-9 anoA-I 1 1 8 2 5 1 3 6 f 8 131+ 22 138f 9’ 1 2 4 f 2 0 1 2 1 f 5 143-C 25 132+ 15 1 4 6 f 12‘ 140% 13

6 5 TG 55+- 4 90’3‘ 9 9 k 4 ’ 85f 19‘ 6 6 2 5‘ 53f 5 81 rL6’ 87 f 11.9* 8 6 k 5’ 7 2 - t - 9

- x- -~ ~

apoA-I1 40 2 2.2 45 f 3.5 49 f 9.1 51 f 4.3’ 46 f 7 39 f 3 39 -C 6.3 40 f 0.5 46 & 2.5 44 f 2.9 The n values indicate the number of experiments at each time point; six subjects were studied with both fat meals.

bp < 0.01 Indicates mean value is significantly greater than corresponding fasting d u e , p < 0.05.

proach to investigating this problem in humans, we have used density gradient ultracentrifugation, polyacrylamide gradient gel electrophoresis, and agarose gel chromatography to ana- lyze changes in the major HDL subclasses during alimentary lipemia. Our results show that after a fatty meal there are increases in concentration of phospholipids, apoA-I, apoA-11, and cholesterol in HDL. There is a major redistribution of HDL mass into subclasses of lower density, without marked alteration in the composition or size of HDL particles in individual sublcasses.

EXPERIMENTAL PROCEDURES~

RESULTS

The plasma lipids, apoA-I and apoA-I1 are shown in Table I. Five h after ingestion of corn oil or corn oil with eggs, there were maximum increases in plasma triglyceride. There were no significant changes in plasma cholesterol at any time point. Six h after fat ingestion, there were maximum increases in plasma apoA-I and apoA-11. These increases were statistically significant ( p < 0.05) for apoA-I and apoA-I1 after ingestion of corn oil and for apoA-I after ingestion of corn oil with eggs. Although the increase in apoA-I1 was smaller after corn oil with eggs, there was no significant difference in the change in apoA-11, comparing the two fat meals.

To analyze the changes within HDL, plasma was analyzed by isopycnic density gradient ultracentrifugation. Density gra- dients between 1.07 to 1.15 g/ml showed the presence of two broad peaks of HDL, shown as distributions of total lipopro- tein, total protein (33), phospholipids, and cholesterol esters in Fig. 1. The distributions of apoA-I and apoA-I1 and the apoA-I/apoA-I1 ratio are shown for the same gradient frac- tions in Fig. 2. In fasting plasma, most of the mass of HDL3 was present in a broad peak of density approximately 1.11 to 1.15 g / d ; a smaller peak representing HDL2, was resolved between density 1.07 to 1.11 g/ml (Figs. 1 and 2, Oh). After corn oil ingestion, the major peak of plasma HDL was in- creased in amount and shifted to lower density; lipoproteins of density 1.07 to 1.11 g/ml also became more prominent and were shifted to slightly lower density (Figs. 1 and 2,Sh). After ingestion of corn oil with eggs, there were similar changes in distribution of total lipoprotein, phospholipid, cholesterol es- ter, and protein and also of apoA-I and apoA-I1 (not shown). The accumulation of lipoprotein components near the bottom of the density gradients (Figs. 1 and 2) suggested that the

* Portions of this paper (including “Experimental Procedures”), are presented in miniprint at the end of this paper. Miniprint is easily read with the aid of a standard magnifying glass. Full size photocopies are available from the Journal of Biological Chemistry, 9650 Rockville Pike, Bethesda, MD 20814. Request Document No. 81M-1411, cite authors, and include a check for $3.20 per set of photocopies. Full size photocopies are also included in the microfilm edition of the Journal that i s available from Waverly Press.

1.14 1.12 1.10 I .08 h

>- I- TI Z

1.16 W 1.14

IO I5

FRACTION NUMBER FIG. 1. Density gradient analysis of fasting plasma (0 h) and

of lipemic plasma (6 h) after ingestion of 100 ml of corn oil. The concentrations of total lipoprotein (W), protein (e), cholesterol ester (O), and phospholipids (O), and the density (A) of individual gradient fractions are shown. Total lipoprotein concentration was estimated from the sum of measured constituents.

more dense HDL was not present in the body of the gradient, as would be expected from the lower density limit of 1.15 g/ ml.

To define in greater detail the changes in the major HDL,? peak, fasting and lipemic plasma were also analyzed by density gradient ultracentrifugation between 1.10 to 1.18 g/ml. An example of the distribution of lipids, total protein, and total lipoprotein mass is shown in Fig. 3. With this type of gradient, the major peak of HDL was well separated from the bottom of the density gradient, but the HDLz (1.08 to 1.10 g / d ) was merged with lower density lipoproteins in the top gradient fractions. In fasting plasma, the major HDL peak was broad, extending from about 1.11 to 1.17 g/ml. A small lipoprotein peak containing phospholipids and apoA-I and unesterified cholesterol (not shown) was present near the bottom of the gradients. After fat ingestion there was an increase in lipopro- tein mass and movement of the center of mass of the main HDL peak toward lower density (Fig. 3 ) . The top of the HDL peak moved from a density of 1.14 g/ml in fasting plasma to

by guest on January 3, 2020http://w

ww

.jbc.org/D

ownloaded from

200 HDL Subclasses after Fat Intake

a density of 1.13 g/ml 4 h after fat ingestion and then to a density of 1.12 g/ml at 6 h. At 8 h, the center of mass of the major HDL peak was returning toward that of fasting HDL (not shown). The small lipoprotein peak near the bottom of the density gradient was more prominent in lipemic samples. Similar changes were observed following ingestion of corn oil with eggs (Fig. 3) or corn oil alone.

In Fig. 4 are shown the mean f S.E. changes in HDL mass as a function of density a t different times after intake of corn oil (circles) or corn oil with eggs (diamonds). To derive these data, in each experiment, the profile of total measured lipo- protein constituents was plotted as a function of density and points along the fasting profile were subtracted from isopycnic points of the lipemic profie. Since the density gradient profides represent near-equilibrium mass distributions, results are shown for material analyzed by either of the two types of density gradient using data from 1.070 to 1.135 g/ml (type of gradient shown in Fig. 1) or 1.100 to 1.170 g/ml (gradient of Fig. 3). In the region of density overlap (1.100 to 1.135 g/ml), similar changes in lipoprotein mass were observed by both types of density gradient technique (Figs. 1 and 3) and thus the data have been averaged.

For both types of fat meal there was a major increase in concentration of lipoproteins of density -1.10 to -1.13 g/ml and a smaller increase in concentration of a separate peak of lipoproteins of density -1.08 to -1.10 g/ml (Fig. 4). The greatest increment occurred in a fairly narrow density interval of 1.110 to 1.125 g/ml. The increments were greatest at 6 h. Also, there was a decrease in concentration of lipoproteins of density -1.14 to -1.17 g/ml. For both types of fat meal, the increments in 1.10 to 1.13 lipoproteins were statistically sig- nificant ( p < 0.05) at all time points for at least two data

FRACTION NUMBER 5 IO I5

rn *

L

5 I O 15

FRACTION NUMBER FIG. 3. Density gradient analysis of fasting plasma (0 h) and

of lipemic plasma 4 and 6 h after ingestion of 80 ml of corn oil with four eggs. Total lipoprotein (m), protein (e), cholesterol ester (O), and phospholipid (0) concentrations and densities (A) are shown for individual gradient fractions.

points within the peak. Similarly, the increments in 1.08 to 1.10 lipoproteins were significant for corn oil at 6 h and for corn oil with eggs at 6 and 8 h. The decrements in 1.137 to 1.165 lipoproteins were significant at 4 h for corn oil with eggs and at 5 and 8 h for corn oil. Although corn oil with eggs caused somewhat greater mean increments of HDL than corn oil alone, these differences were not statistically significant.

The changes in individual HDL constituents, expressed as milligrams per dl of plasma, were determined for the density intervals defined by the major peaks and troughs of the difference gradients of Fig. 4. There were no significant differ- ences in response to corn oil or corn oil with eggs, whether compared a t individual time points or over all time points. Thus, the results for both types of fat meal have been com- bined and are shown as changes in lipid concentration in Fig. 5 and as changes in apoA-I and apoA-I1 concentration in Fig. 6. The increase in 1.080 to 1.102 lipoproteins was largely due to increased concentrations of phospholipid, cholesterol ester, and apoA-I. The increase in 1.102 to 1.137 lipoproteins was due to increased concentrations of apoA-I1 and cholesterol as well as apoA-I, phospholipids, and cholesterol esters. The decrements in 1.137 to 1.165 lipoproteins were due to de- creased concentrations of cholesterol esters, and possibly also an0A-T and moA-11. The statistical significances of these

FIG. 2. Density gradient analysis of fasting (0 h) and lipemic (6 h) plasma, showing distribution of apoA-I (0) and apoA-11 0 and the apoA/apoA-LI ratio (O), as determined by radioim- munoassay. The concentrations are given as milligrams of protein per ml of density gradient fraction. These are the same density gradient fractions as shown in Fig. 1. -r--- - - ~ ~ - - - 1~ ~-~

by guest on January 3, 2020http://w

ww

.jbc.org/D

ownloaded from

HDL Subclasses after Fat Intake 201

changes are given in the legends to Figs. 5 and 6. The molecular weights (M,) of lipoproteins isolated in dif-

ferent density intervals are shown in Table 11. In order to evaluate changes in composition and molecular weight across the density spectrum of HDL, lipoproteins were analyzed in narrow density intervals. Going across the density intervals from 1.150 to 1.110 g/ml lipoproteins showed small, progres- sive increases in molecular weight, whereas the 1.080 to 1.095 lipoproteins were considerably larger. The molecular weights of HDL subfractions in lipemic plasma were similar to those in the same density interval of fasting plasma.

To compare HDL species with those reported (16, 19, 34) sedimentation coefficients were determined on selected sam- ples from the density gradients. The densities, molecular weights, and sedimentation coefficients of fasting HDL are shown in Table I11 and compared with those reported by Anderson et al. (19) and Patsch et al. (16). Apart from a somewhat higher molecular weight of the least dense HDL

I .or L I A

z a W

z a 0 X

1.06 1.08 1.10 1.12 1.14 1.16 1.18

DENSITY fg/ml) FIG. 4. Changes in concentration of total HDL constituents

at different time points after fat ingestion. For each experiment, the change in lipoprotein concentration was measured by subtracting the density gradient profile of fasting HDL from that of lipemic HDL. The difference in concentration was obtained at each 0.005 g / d interval along the curves of total lipoprotein concentration. The mean f S.E. changes are shown following ingestion of corn oil (@) or corn oil with eggs (e). At each time point, results are shown for four or five meals of corn oil (in six subjects) or four meals of corn oil with eggs (in the same six subjects).

obtained by Anderson et al. (19), the molecular weights and sedimentation coefficients of the different HDL subfractions are in close agreement in the three studies. The molecular weights of the 1.080 to 1.095 and 1.135 to 1.150 lipoproteins

L I I

4 5 6 7 8

TIME (h) FIG. 5. Change in concentration of HDL lipids in different

density intervals after fat ingestion. The values shown are the mean f S.E. changes expressed per dl of plasma, for phospholipids (0), cholesterol ester (0). and cholesterol (e). The density intervals were chosen to match the peaks and troughs of the difference gra- dients of Fig. 4. At each time point, n = 8 or 9. Lipid recoveries were 90 to 95%. The statistically significant changes were: 1.080 t.o 1.102 phospholipids at 5, 6, and 8 h ( 2 p < 0.05, 0.001, 0.01), cholesterol esters at 5.6, and 8 h (2 p < 0.05), cholesterol at 6 and 8 h (2 p < 0.05, 0.01); 1.102 to 1.137, phospholipids at 4,5,6, and 8 h (2 p < 0.05,0.005, 0.01, 0.001) cholesterol esters at 5 and 6 h ( 2 p < 0.05), cholesterol at 6 h (2 p e 0.05) (the change was 0.63 -C 0.2 mg/dl); 1.137 to 1.165, cholesterol esters at 5 and 6 h (2 p < 0.025, 0.05).

I O -

2 P

V * r 10-

- - - - - - - - - - - - - - - -. -lo -

4 5 6 7 8

TIME (h) FIG. 6. Mean f S.E. changes in concentration of HDL apo-

proteins (milligrams per plasma) in different density intervals after fat ingestion. At each time point n = 8 or 9. Recoveries were 80 to 85% for apoA-I and 90 to 95% for apoA-11. The statistically significant changes were: 1.080 to 1.102, apoA-I at 6 h ( 2 p < 0.05); 1.102 to 1.137, apoA-I at 5, 6, and 8 h ( 2 p < 0.05), apoA-I1 at 4, 5, 6, and 8 h ( 2 p < 0.01, 0.001, 0.001. 0.01); 1.137 to 1.165, apoA-11 at 4 h (2 p < 0.05). Data Were not available for the 8-h time point in the 1.137 to 1.165 g/ml density interval.

by guest on January 3, 2020http://w

ww

.jbc.org/D

ownloaded from

202 HDL Subclasses after Fat Intake

TABLE I1 Molecular weight and composition of HDL subfractions

Table shows mean f S.E. compositional values obtained from analysis of individual fractions in the density regions shown. Total protein was estimated by the Lowry method and apoA-I and apoA-I1 by radioimmunoassay. Molecular weights are mean values (n = 2) or mean f S.E. (n = 3 to 7) values for pooled fractions in the density regions shown.

Density W m l )

1.080-1.095 1.110-1.120 1.120-1.135 1.135-1.150

Basal M, X 378 f 6 248 f 12 219 ApoA-I/ApoA-II 4.7 f 6 2.2 f 0.4 1.9 f 0.2 %pro" 46.1 f 3.6 53.6 f 2.6 58.0 f 1.2 %PL 23.5 f 2.3 22.0 f 1.2 19.1 f 1.0 %C 5.0 f 0.7 2.6 f 0.31 %CE

2.1 If: 0.2 23.0 f 2.2 19.6 f 1.5 18.8 f 0.8

%TG 2.1 f 0.4 1.5 f 0.2 1.5 f 0.1 4 h

ApoA-I/ApoA-II 5.4 f 1.6 1.9 f 0.2 2.0 f 0.4 %pro 42.5 f 4.5 49.9 f 2.8 56 f 2.1 %PL 28.0 f 3.7 25.0 f 2.4 22.4 f 2.0 %C 4.9 f 0.5 3.1 f 0.5 2.0 f 0.2 %CE 23.0 f 0.6 20.2 f 1.2 18.1 f 0.6 %TG 1.9 f 0.4 1.6 f 0.3 1.6 f 0.2

5 h M , X 250 216 ApoA-I/ApoA-II 5.1 f 1.2 2.0 f 0.3 2.0 f 0.6 %pro 43.6 f 3.5 53.4 f 1.4 57.0 f 1.0 %PL 27.2 f 2.2 25.2 f 1.1 21.6 f 0.8 %C 4.6 f 0.5 2.0 f 0.4 1.9 f 0.1 %CE 22.5 f 1.8 17.6 f 0.7 17.4 f 0.4 %TG 2.2 f 0.3 1.5 f 0.3 1.7 f 0.3

M , x 386 244 f 5 212 ApoA-I/ApoA-II 4.6 f 1.4 2.1 f 0.4 2.0 f 0.5 %pro 37.0 f 5.1 47.8 f 2.7 54.5 f 2.9 %PL 32.0 f 3.2 28.2 f 2.5 23.7 f 2.8 %C 4.6 f 0.8 2.7 f 0.3 1.9 f 0.2 %CE 24.2 f 1.7 19.8 f 1.0 18.2 f 1.0 %TG 2.4 f 0.6 1.7 f 0.2 1.8 f 0.2

6 h

8 h M , x 234 218 ApoA-I/ApoA-II 2.1 f 0.5 2.1 f 0.6 %pro 42.3 f 4.1 52.5 f 2.5 55 f 2.2 %PL 28.1 f 2.4 24.2 f 1.6 23.8 f 1.6 %C 4.9 f 0.4 2.7 f 0.2 1.8 f 0.1 %CE 22.5 f 1.8 19.0 f 1.2 17.5 f 0.8 %TG 2.3 f 0.4 1.7 f 0.3 1.6 f 0.3

The abbreviations used are: pro, protein; PL, phospholipid; C, cholesterol; CE, cholesterol ester; TG, triglyceride.

173 f 3 3.0 f 0.5

60.1 f 1.4 18.3 f 0.9

18.9 f 1.0 1.7 f 0.11

1.0 f 0.1

59.5 22.7

1.5 15.2 1.1

176 3.1 f 0.7

65.0 18.1 1.7

14.3 0.9

167 3.2 f 0.5

60.0 22.4

2.4 14.0 1.2

179 3.2 f 0.4

61.5 22.1 1.8

13.8 0.8

also agree with values reported by Shen et al. (34) for HDL species of similar density. In our study, identical sedimenta- tion coefficients were obtained for HDL subclasses isolated from the same density interval of fasting and lipemic plasma.

Compositional analysis (Table 11) showed that the 1.080 to 1.095 lipoproteins had a high apoA-I/apoA-I1 ratio, while the 1.110 to 1.135 lipoproteins were distinguished by their low ratio of apoA-I to apoA-11. Lipoproteins of d 1.135 to 1.150 had somewhat higher ratios of apoA-I to apoA-I1 than the 1.110 to 1.135 lipoproteins. During lipemia the apoA-I/apoA- I1 ratios were remarkably constant within individual density cuts. The sum of apoA-I and apoA-I1 masses was equal to about 98% of the total protein (33) of lipoproteins of d > 1.11 and 94 to 96% of that of HDL of d < 1.11. Isoelectric focusing of HDL apoproteins showed that apoproteins E and C were minor constituents of d > 1.11 lipoproteins in both fasting and lipemic plasma. During lipemia, all lipoproteins showed an increased per cent phospholipid; the 1.080 to 1.095 and 1.110 to 1.135 lipoproteins showed a corresponding decreased per cent protein while the 1.135 to 1.150 lipoproteins showed a decreased per cent cholesterol ester.

The average molecular compositions of HDL within the different density intervals were calculated from their mean compositions and mean molecular weights (Table IV). Except

for a small increase in number of phospholipid molecules, HDL particles in lipemic plasma has a similar average molec- ular composition to those in fasting plasma. The 1.080 to 1.095 and 1.135 to 1.150 fractions did not show integral numbers of apoprotein molecules per particle, indicating that these frac- tions did not contain single species of particles, The 1.110 to 1.135 fraction contained particles with an apoprotein compo- sition approaching 3 molecules of apoA-I and 2 molecules of apoA-11, suggesting a fairly homogeneous population. Fur- thermore, the apoprotein content of particles across the 1.110 to 1.135 density range was constant, whereas the particles had an increased content of lipid in the lower density interval (1.110 to 1.120).

To assess further changes in the distribution of HDL during lipemia, holo-HDL (d 1.063 to 1.210 g/ml) was prepared from fasting and lipemic plasma of four subjects (two males and two females) and analyzed by polyacrylamide gradient gel electrophoresis. This technique separates lipoprotein particles on the basis of size and has been shown to resolve four or five partially separated bands when applied to fasting human HDL (19-21). For fasting HDL, all subjects displayed a major band of HDL, corresponding to a particle diameter of 9.1 to 9.4 nm, two smaller bands of diameters 8.3 to 8.6 nm and 7.9 to 8.3 nm, and one or two larger bands, a well resolved band of

by guest on January 3, 2020http://w

ww

.jbc.org/D

ownloaded from

HDL Subclasses after Fat Intake 203

TABLE I11 Molecular weights of HDL subfractions

This study Anderson et al. (19) Patsch el al. (16)

N Lipoprotein Mr (x Lipoprotein Mr (x 10-3) s;l Lipopro- S? density density k i n sity den- (x'& s:

g/ml g/ml g/ml 3 1.080-1.095 378 f 6" 5.75 1.063-1.10 410-429 5.4 1.096 358 5.3*

(1.09) 7 1.11-1.12 248 * 12 3.5 1.10-1.125 262-264 3.2

(1.11) 4 1.135-1.15 173 It 3 1.55 1.125-1.20 177 1.56 1.143 182 1.52

(1.145)

a Standard error for n determinations performed on samples from two fat-feeding studies. ' Corrected to density 1.20 from 1.21 g/ml to correspond to the other two groups. S;? is in density 1.20 g/ml of NaBr-NaC1 solvent of

Anderson et al. (19).

TABLE IV Average molecular composition of HDL

Number of molecules per particle calculated from mean per cent compositions and molecular weights. Proteins other than apoA-I and apo1-I1 are assumed to represent 4% of the proteins of HDL2 and 2% of those HDL.

Density (g/ml)

1.080-1.095 1.110-1.120 1.120-1.135 1.135-1.150

Basal ANA-I APOA-I1 PL" C CE TG

5 h APOA-I AeoA-I1 P i C CE TG

6 h APOA-I APOA-I1 PL C CE TG

8 h APOA-I

PL ApoA-I1

C CE TG

4.8 3.2 2.9 2.7 1.8 2.2 2.3 1.4

114 70 55 41 49 17 12 8

130 73 62 49 9 5 4 2

3.0 2.8 3.0 2.3 2.3 1.4

81 60 41 13 11 8 65 56 38 4 4 2

4.2 2.8 2.7 2.6 1.5 2.1 2.2 1.3

159 89 65 48 46 17 10 10

140 72 58 35 10 5 4 2

2.9 2.8 2.9 2.1 2.2 1.4

72 67 51 16 10 8 65 58 37

5 4 2 The abbreviations used are: PL, phospholipid; C, cholesterol; CE,

cholesterol ester; TG, triglyceride.

diameter 10.5 to 10.8 nm and a variable, minor one of diameter 11.2 to 11.8 nm. The female subjects showed relatively greater amounts of these larger HDL species. During lipemia, a similar pattern of change was observed in all subjects. No major new band of HDL was evident. However, there was a redistribution of HDL mass, with increased prominence of the two major HDL species and decreased prominence of the two smallest HDL species. During lipemia, each band of HDL appeared to be shifted to a slightly larger particle size. This was especially evident for the major HDL species which was shifted from a peak of 9.1 to 9.4 nm in fasting HDL to 9.4 to 9.8 nm in lipemic HDL. Gradient gel electrophoresis of HDL obtained 4,6,8, or 10 h after fat ingestion showed that the maximum changes were observed at 6 to 8 h with a return to the fasting pattern at 10 h. Identical patterns were obtained upon repeated ex- amination of the same samples, with less than 0.1 nm variation

in peak sizes. In further experiments designed to compare the results of

density gradient and gradient gel analysis, gradient gels were performed on individual density gradient fractions from fast- ing (Fig. 7A) and lipemic (Fig. 7B) plasma. The density gradient was similar to that shown in Fig. 3. The fractions near the top of the density gradients (d - 1.103 g/ml) showed multiple bands, as did fractions of density -1.14 g/ml or greater. Between d - 1.11 to d - 1.13 g/ml, there was a single band of HDL which showed a decrease in size from about 10.0 to 9.0 nm with decreasing density. Similar sized particles were isolated from fractions of similar density for fasting and li- pemic plasma. For the experiment shown, the greatest mass of lipoprotein was present at d 1.137 g/ml in fasting plasma and at d 1.121 g/ml in lipemic plasma, indicating that the major species of HDL had a diameter of -9.1 nm in fasting plasma and of -9.7 nm in lipemic plasma. Similar experiments performed using gradients such as shown in Fig. 1 showed that the HDL peak of density -1.080 to 1.100 g/ml contained a single band of diameter 10.5 to 11.0 nm, in both fasting and lipemic plasma. These results indicate that the two major incremental species of densities -1.12 and -1.09 g/ml (Fig. 4) correspond, respectively, to the 9.4 to 9.8 and 10.5 to 10.8 nm bands shown by gradient gel electrophoresis of lipemic HDL. Moreover, the molecular weights of the HDL species isolated by density gradient indicate particle diameters corresponding to those observed by gradient gel electrophoresis."

To assess the possible role of prolonged ultracentrifugation in producing changes in HDL distribution, plasma was also analyzed by brief centrifugation (4 h) in the vertical rotor (Fig. 8). The resulting density gradients and the distribution of cholesterol, apoA-I, apoA-11, and apoA-I/apoA-I1 ratios are shown before (Fig. 8, top panels) and after (Fig. 8, bottom panels) ingestion of corn oil. In fasting plasma there was a single major peak of HDL isolated between densities 1.11 to 1.20 g/ml and a poorly defined lesser peak of density approx- imately 1.08 to 1.11 g/ml. The major change during lipemia was an increase in apoA-I, apoA-11, cholesterol, and phospho- lipids in the less dense part of the major HDL peak (d 1.11 to 1.14 g/ml) and a small decrease in mass of apoA-I, apoA-11,

V - MJdN, where V = volume of one HDL particle, M, = molecular weight, N = Avogradro's number, d = equilibrium banding density. Thus,

.-( ?rdN ] 8 M , x 10" ,,3

where D is the paticle diameter in nanometers, i.e. D - (.00423 M,/ For the major incremental HDL peaks in lipemic plasma M, =

244,000, d = 1.115 g/ml:. D = 9.7 nm; and M, = 386,000, d = 1.095 g/ ml ;. D = 11.4 nm. These values are similar to the sizes of the two major HDL species determined by gradient gel electrophoresis.

by guest on January 3, 2020http://w

ww

.jbc.org/D

ownloaded from

HDL Subclasses after Fat Intake 204

W V z 3 5: m a

B - 10.0

1.104 I I

8.9 8.6 - 1.140 I

1.189

MIGRATION - FIG. 7. Gradient gel scans of individual density gradient

(SW 50.1) fractions obtained from fasting plasma (A) and lipemic plasma (6 h, B) in the HDL region. The densities (grams per ml) of the fractions are shown to the left and the lipoprotein diameter (nanometers) is shown above the peaks. All fractions be- tween 1.10 to 1.14 g/ml have been shown; representative fractions are shown for the 1.14 to 1.18 g/ml density region. The density gradient was similar to that shown in Fig. 4.

and cholesterol in the more dense (>1.14 g/ml) HDL fractions. Similar results were obtained in three separate experiments. It is notable that a peak of apoA-I and phospholipids was not observed at or near the bottom of the vertical rotor density gradients, as was noted in the gradients prepared by SW 50.1 ultracentrifugation (Fig. 3), suggesting that this peak arose as a result of dissociation of these components from HDL parti- cles during prolonged ultracentrifugation. However, a similar distribution of apoA-I, apoA-11, and lipids was obtained by both the SW 50.1 and vertical rotor techniques, with a nadir in apoA-I/apoA-I1 ratio at density 1.11 to 1.14 g/ml. This shows that ultracentrifugation did not cause a major redistri- bution of apoA-I and apoA-I1 mass between HDL species.

Gradient gel electrophoresis was performed on vertical rotor fractions of d 1.08 to 1.10 g/ml, d 1.10 to 1.13 g/ml, and d 1.13 to 1.15 g/ml during fasting and lipemia. Similar sized particles were isolated from each density interval during fasting and lipemia. The 1.08 to 1.10 fraction contained a single band of diameter 11.0 nm. The 1.10 to 1.13 fraction contained two bands of diameters 10.7 and 9.8 nm, and the 1.13 to 1.15 fraction contained bands of diameters 9.3, 8.7, and 8.2 nm. Thus, the major increment in HDL mass occurred in the density interval containing 10.7 and 9.8 nm HDL species. These results are similar to those obtained by much longer ultracentrifugation in the SW 50.1 rotor except that the two larger species of HDL were not well separated from each other by the vertical rotor technique. Also, the vertical rotor frac- tions were heavily contaminated with albumin.

Fasting and lipemic plasma were also analyzed by chroma- tography on 6% agarose. The distribution of apoA-I and apoA-

I1 is shown in Fig. 9A and that of phospholipids and total cholesterol in Fig. 9B. Agarose chromatography resolved only a single peak of HDL. However, during lipemia (Fig. 9, dashed lines) all components of the HDL peak were shifted to a lower elution volume. There was an increase in mass of apoA-I, apoA-11, cholesterol, and phospholipids in larger HDL species eluting in the left-hand side of the peak and a decrease in mass of apoA-I, apoA-11, phospholipids, and cholesterol elut- ing on the right-hand side of the peak. By contrast, the elution

2 0 c oh

I 5 - IO -

5 -

0.4

0.2

FRACTION NUMBER FIG. 8. Vertical rotor density gradient analysis of fasting

(0 h) and Lipemic (6 h) plasma. The first and third panels show the density (A) and apoA-I/apoA-I1 weight ratios (0) of the gradient fractions. The second and fourth panels show the concentrations of apoA-I (O), apoA-I1 (O), and total cholesterol (W).

by guest on January 3, 2020http://w

ww

.jbc.org/D

ownloaded from

HDL Subclasses after Fat Intake 205

ELUTION VOLUME (ml) FIG. 9. Agarose gel chromatography of fasting (-) and 6-

h lipemic (- - -) plasma, showing the HDL region of the column elution profile. The concentrations (micrograms per ml of column fraction) of apoA-I (A, A) and apoA-I1 (0 ,O) are shown in the top panel and of phospholipid (A, A) and total cholesterol (0, 0) in the bottom panel. Two ml plasma samples were chromatographed.

volume of the LDL peak (V, = 114 m l ) and the void (V, = 72 m l ) and salt (V , = 190 m l ) peaks were unchanged during lipemia. No free apoA-I was detected by column chromatog- raphy. Pooled fractions across the agarose column profile were centrifuged at d 1.21 g/ml for 36 h and the supernatant fractions were examined by gradient gel electrophoresis. Frac- tions of elution volume 136 to 142 ml contained a single band of diameter 10.7 (fasting) and 11.0 nm (lipemic); fractions from 143 to 147 ml contained two bands of diameters 10.5 nm (fasting and lipemic); fractions of elution volume 148 to 152 ml contained three bands of diameters 9.1 (major), 8.3, and 7.8 nm (fasting and lipemic); and fractions of elution volume 153 to 159 ml contained three bands of diameters 9.0, 8.3 (major), and 7.6 nm (fasting and lipemic). Thus, as in the centrifugal experiments, the major increase in HDL mass occurred in fractions containing two separate species of HDL, of diameters 10.5 and 9.5 nm. These results show a similar redistribution of HDL mass by a noncentrifugal technique and indicate that the larger HDL species were not generated from smaller HDL species during ultracentrifugation.

DISCUSSION

Our findings c o n f i i the earlier observations showing an increase in HDL phospholipids and proteins during alimen- tary lipemia (14, 15). Development of immunoassays (35, 36) has allowed measurement of specific apoproteins in whole plasma and in column or density gradient fractions. By use of radioimmunoassays, we were able to show that the increased protein mass in HDL was associated with increments in apoA- I and, to a lesser extent, apoA-11. In view of the evidence that phospholipids (7, 13) and apoA-I (7) are transferred from chylomicrons into HDL in the rat, it is likely that the increase in mass of HDL phospholipids, apoA-I, and apoA-I1 reflects a similar process in humans. However, it should be noted that direct evidence for transfer of chylomicron surface constitu-

ents into HDL was not obtained in the study. The similarity of changes in the HDL following ingestion of corn oil or corn oil with eggs suggests that acute ingestion of cholesterol with triglyceride does not alter the contribution of chylomicron surface constituents to plasma HDL.

Our principal aim was to study the detailed changes in composition, size, and density distribution of HDL during lipemia. In general, a similar pattern of change in the plasma HDL was recorded by agarose chromatography and by density gradient ultracentrifugation in the swinging bucket or vertical rotors. Chromatography showed that lipemia was associated with a redistribution of HDL with an increase in mass in the larger particles and a decrease in mass in the smaller particles of the HDL peak (Fig. 9). Vertical rotor (Fig. 8) or swinging bucket (Figs. 1 to 3) density gradient ultracentrifugation of lipemic plasma also showed an increase in mass of the larger, less dense HDL, and a decrease in mass of the more dense HDL. This pattern of change suggests that transfer of chylo- micron surface constituents into HDL was associated with conversion of smaller HDL particles into larger HDL particles, probably due to insertion of chylomicron surface constituents into pre-existing HDL species. An analogous conversion of HDLs into larger, less dense particles has also been observed following heparin-stimulated lipolysis in hypertriglyceridemic subjects (37), and as a result of in vitro lipolysis of very low density lipoprotein in the presence of HDL, (1 1).

More detailed analysis indicated that the increment in larger, less dense HDL was due to increased amounts of two major, separate HDL species. Calculation of the lipemic-fast- ing differences in HDL mass distribution (Fig. 4) revealed increments of two separate components of densities approxi- mately 1.09 and 1.12 g/ml. Similarly, gradient gel electropho- resis of holo-HDL showed increases in two major bands of larger HDL, of diameters 10.5 to 10.8 nm and 9.4 to 9.8 nm. Gradient gel electrophoresis of individual density gradients fractions showed that HDL species of these diameters were isolated in the density intervals 1.080 to 1.100 g/ml and 1.100 to 1.125 g/ml, indicating equivalence of the two incremental species observed by the different techniques. Gradient gel electrophoresis of HDL fractionated by vertical rotor or aga- rose column also showed that regions of greatest increase in lipoprotein mass contained major species of diameters ap- proximately 10.8 nm and 9.5 to 9.8 nm. Thus, although pro- longed ultracentrifugation was associated with dissociation of some HDL components, centrifugation per se did not create the lipemic pattern of increment in two separate large species of HDL.

Using mean molecular weights and per cent compositions, we were able to estimate average molecular compositions for particles in the different density fractions of HDL (Table IV). It is notable that particles across the 1.110 to 1.135 g/ml density interval had an invariant apoprotein complement, approaching 3 molecules of apoA-I and 2 molecules of apoA- I1 per particle. Thus, the variation in size and molecular weight across this density interval was entirely due to enrich- ment with lipid. In fasting plasma, the major component in the 1.120 to 1.135 density interval was a particle with a diameter of -9.2 nm (Fig. 7 ) , corresponding to the major component of fasting HDLB (19-21). During lipemia, the most intense increment occurred in HDL of d 1.110 to 1.120 g/ml. Relative to the fasting 1.120 to 1.135 fraction, particles in the 6-h lipemic a! 1.110 to 1.120 fraction contained about 62% more phospholipid molecules and 41% more cholesterol molecules with smaller increases in cholesterol ester (16%) and triglyc- eride (25%).

These results suggest that the increase in 1.1 1 to 1.12 lipoproteins during lipemia is due to addition of lipid mole-

by guest on January 3, 2020http://w

ww

.jbc.org/D

ownloaded from

206 HDL Subclasses after Fat Intake

cules to the major species of fasting HDLa. The lipid enrich- ment process could be initiated by insertion of chylomicron phospholipids into pre-existing HDL particles. Cholesterol might be transferred into the phospholipid-enriched lipopro- tein surface from other lipoproteins or from cell membranes, and small amounts of cholesterol ester might accumulate as a result of the action of 1ecithin:cholesterol acyltransferase. Conversion of HDL3 into a slightly larger particle of d 1.11 to 1.12 g/ml is the dominant mechanism of phospholipid uptake by HDL when egg phosphatidylcholine vesicles are incubated with isolated HDL3 or the plasma (38). In plasma, the phos- pholipid enrichment of HDL is accompanied by influx of unesterified cholesterol from other lipoproteins (38). Omission of 1ecithin:cholesterol acyltransferase inhibitor from these in- cubations results in formation of slightly larger HDL particles, of identical size to the HDL observed in the 1.11 to 1.12 interval during l i~emia .~ Although compositional data can only provide circumstantial evidence for the metabolic origins of lipoproteins, these reconstitution studies also suggest that d 1.11 to 1.12 HDL may be derived by lipid enrichment of

Within density intervals, the molecular weights, diameters, sedimentation coefficients, and mean compositions of HDL did not vary greatly between fasting and lipemic plasma. Thus, lipemia does not cause appearance of major new species of HDL, but rather a redistribution of HDL mass between existing species. The similar particle contents of apoA-I and apoA-11 in fasting and lipemic plasma indicate that apoprotein transfer into HDL is not associated with formation of major new species of novel apoprotein composition. Thus, apopro- tein transfer may cause interconversion of existing HDL spe- cies or may result in formation of new particles with the same apoA-I/apoA-I1 ratio as existing species. Scanu and co-work- ers have shown that apoA-I1 but not apoA-I can be inserted into HDL3 in vitro (39, 40). It is of interest to note that the 1.110 to 1.135 lipoproteins have approximately one more apoA-I1 molecule than the 1.135 to 1.150 lipoproteins (Table IV). Thus, transfer of apoA-I1 into HDL might involve con- version of smaller HDLa species with a high apoA-I/apoA-I1 ratio into larger species containing an extra molecule of apoA- 11. This could explain why lipid enrichment of the larger HDLa was not accompanied by depletion of HDL in the d 1.120 to 1.135 density interval.

The results of the present study can be compared with earlier investigations of the heterogeneity of HDL using fixed angle (19), swinging bucket (41) or zonal rotor (16) density gradient techniques. Anderson et al. (19) reported that a peak of HDL of d 1.100 to 1.125 g/ml, designated “HDLz,,” could be well resolved from the major peak of HDLs by density gradient ultracentrifugation. In subsequent studies (16, 41) including the present one, a separate peak of HDLz, was not resolved by density gradient ultracentrifugation. However, all investigators (16, 19, 41) found that a significant part of the HDL3 peak extends into the 1.100 to 1.125 g/ml region, and in the present study, this material was shown to be recovered in the same density interval upon recentrifugation. Furthermore, this density interval contained particles of similar Stokes radius, molecular weight, and sedimentation coefficient to the HDLz, described by Anderson et al. (19). Thus, although we have been unable to isolate HDL2. as a separate peak, the HDL of d 1.11 to 1.12 g/ml of the present study have similar properties to HDL2.. Our results suggest that HDL of d 1.11 to 1.12 g/ml may not be a separate subclass of HDLz, but rather a lipid-enriched form of the major species of HDL3. These particles appear to accumulate during conditions of stimulated lipid transfer into HDL.

HDL3.

A. Tall, unpublished observations.

Acknowledgments-We wish to acknowledge the excellent tech- nical assistance of Edith Abreu and Lynne Robinson.

REFERENCES 1. Rhoads, G. G., Gulbrandsen, C. L., and Kagan, A. (1976) N. Engl.

J. Med. 294,293-298 2. Gordon, T., Castelli, W. P., Hjortland, M. C., Kannel, W. B., and

Dawber, T. R. (1977) Am. J. Med. 62, 707-714 3. Miller, G. J., and Miller, N. E. (1975) Lancet 1, 16-19 4. Hamilton, R. L., Williams, M. C., Fielding, C. J., and Havel, R. J.

5. Green, P. H. R., Tall, A. R., and Glickman, R. M. (1978) J. Clin.

6. Forte, T. M., Nichols, A. V., Gong, E. L., Lux, S., and Levy, R. I.

7. Tall, A. R., Green, P. H. R., Glickman, R. M., and Riley, J. W.

8. Glickman, R. M., Green, P. H. R., Lees, R., and Tall, A. R. (1978)

9. Schaefer, E. J., Jenkins, L. L., and Brewer, H. B. (1978) Biochem.

10. Mjos, 0. D., Faergeman, O., Hamilton, R. L., and Havel, R. J.

11. Patsch, J. R., Gotto, A. M., Olivecrona, T., and Eisenberg, S.

12. Chajek, T., and Eisenberg, S. (1978) J. Clin. Znuest. 63,1654-1665 13. Redgrave, T. G., and Small, D. M. (1979) J. Clin. Znuest. 64,162-

14. Havel, R. J. (1957) J. Clin. Znuest. 36,848-854 15. Havel, R. J., Kane, J. P., and Kashyap, M. L. (1973) J. Clin.

16. Patsch, W., Schonfeld, G., Gotto, A. M., Jr., and Patsch, J. R.

17. Weisgraber, K. H., and Mahley, R. W. (1980) J. Lipid Res. 21,

18. Anderson, D. W., Nichols, A. V., Pan, S. S., and Lindgren, F. T. (1978) Atherosclerosis 29, 161-179

19. Anderson, D. W., Nichols, A. V., Forte, T. M., and Lindgren, F. T. (1977) Biochim. Biophys. Acta 493,46-68

20. Nichols, A. V., Blanche, P. J., Krauss, R. M., and Gong, E. L. (1979) Circulation 59-60 Suppl., 11-118

21. Nichols, A. V., Blanche, P. J., Krauss, R. M. and Gong, E. L. (1979) in Report of the HDL Methodology Workshop, Sun Francisco, May, 1979 (Lippel, K., ed) National Institutes of Health Publication 79-1661, July

22. Fielding, P. E., and Fielding C. J. (1980) Proc. Natl. Acad. Sci. U.

23. Chung, B. H., Wilkinson, T., Geer, J. C., and Segrest, J. P. (1980)

24. Folch, J., Lees, M., and Sloane, G. H. (1957) 226,497-509 25. Bartlett, G. R. (1959) J. Biol. Chem. 234,466 26. Zlatkis, A., and Zak, B. (1969) Anal. Biochem. 29, 143 27. Blum, C. B., Aron, L., and Sciacca, R. (1980) J. Clin. Znuest. 66,

28. Lux, S. E., John, K. M., and Brewer, H. B., Jr. (1972) J. Biol.

29. Greenwood, F. C., Hunter, W. M., and Glover, J . S. (1963) Bio-

30. Warnick, C. R., Mayfield, C., Albers, J. J., and Hazzard, W. R.

31. Weber, K., and Osborn, M. (1969) J. Biol. Chem. 244,4406-4412 32. Nelson, C. A,, Lee, J . A,, Brewster, B., and Morris, M. D. (1974)

33. Lowry, 0. H., Rosebrough, N. J., Farr, A. L., and Randall, R. J.

34. Shen, B. W., Scanu, A. M., and Kezdy, F. J. (1977) Proc. Natl.

35. Schonfeld, G., and Pfleger, B. (1974) J. Clin. Znuest. 54,236-246 36. Cheung, M. C., and Albers, J. J . (1977) J. Clin. Znuest. 60,43-50 37. Forte, T. M., Krauss, R. M., Lindgren, F. T., and Nichols, A. V.

38. Tall, A. R., and Green, P. H. R. (1981) J. Bwl. Chem. 256,2035-

39. Scanu, A. M., Lagocki, P., and Chung, J. W. (1980) Ann. N. Y.

40. Lagocki, P. A,, and Scanu, A. M. (1980) J. Biol. Chem. 255,2701-

41. Cheung, M. C., and Albers, J. J. (1979) J. Lipid Res. 20,200-207

(1976) J. Clin. Znuest. 58, 667-680

Znuest. 61,528-534

(1971) Biochim. Biophys. Acta 248, 381-386

(1979) J. Clin. Znuest. 64,977-989

N. Engl. J. Med. 299, 1424-1427

Biophys. Res. Commun. 80,405-412

(1975) J. Clin. Znuest. 56,603-615

(1978) Proc. Natl. Acad. Sei. U. S. A. 75, 4519-4523

171

Znuest. 52,32-38

(1980) J. Biol. Chem. 255,3178-3185

316-325

S. A . 77,3327-3330

J. Lipid Res. 21, 284-291

1240-1250

Chem. 247, 7510-7518

chem. J. 89, 114-123

(1979) Clin. Chem. 25,279-284

Anal. Biochem. 59, 69-74

(1951) J. Biol. Chem. 193,265-275

Acad. Sci. U. S. A. 74,837-841

(1979) Proc. Natl. Acad. Sci. U. S. A . 76, 5932-5938

2044

Acad. Sci. 348, 160-173

2706

by guest on January 3, 2020http://w

ww

.jbc.org/D

ownloaded from

HDL Subclasses after Fat Intake 207

by guest on January 3, 2020http://w

ww

.jbc.org/D

ownloaded from

A R Tall, C B Blum, G P Forester and C A Nelsonafter ingestion of fat.

Changes in the distribution and composition of plasma high density lipoproteins

1982, 257:198-207.J. Biol. Chem. 

  http://www.jbc.org/content/257/1/198Access the most updated version of this article at

 Alerts:

  When a correction for this article is posted• 

When this article is cited• 

to choose from all of JBC's e-mail alertsClick here

  http://www.jbc.org/content/257/1/198.full.html#ref-list-1

This article cites 0 references, 0 of which can be accessed free at

by guest on January 3, 2020http://w

ww

.jbc.org/D

ownloaded from